Two measurements (for the Hudson Bay and Baffin Bay) performed by the Naval Underwater Systems Center to determine volume attenuation at low frequencies have been reanalyzed. The reanalysis gives substantial evidence that the attenuation coefficients previously inferred for these areas are dominated by the influence of range dependence in the respective environments. The previous analysis assumed cylindrical spreading for the long‐range propagation loss, with departures ascribed to volume attenuation. The work here shows that such departures may be caused by range dependence of the environment for the Hudson and Baffin Bays. The conclusion is that noaprioriassumption about the dependence of loss on range is defensible for the long ranges involved in these experiments and that careful concurrent environmental measurements are required for use in the data analysis.